I want to strip the "other" array if its length is 0
Here is my schema
languages: Yup.object({
native: Yup.string().oneOf(languages),
other: Yup.array()
.max(5)
.of(
Yup.object({
language: Yup.string().oneOf(languages),
speaking: Yup.string().oneOf(fluency),
reading: Yup.string().oneOf(fluency),
writing: Yup.string().oneOf(fluency),
})
)
.when("other.length", {
is: 0,
then: (s) => s.strip(),
}),
}),
The error I get:
Uncaught Error: Cyclic dependency, node was: "other"
Thank beforehand
For someone who may encounter the same issue:
I found the issue in:
.when(".length", {
is: 0,
then: (s) => s.strip(),
I needed to reference the .length of the array rather than the arr.length and this resolved the problem.
I have a costly server ajax request which has one input (full: boolean). If full is false, the server can return either a partial or a full response (response.isFull == true); but if full is true, the server will return a full response. Normally the partial response is good enough, but there are certain conditions that will require a full response. I need to avoid requesting a full response explicitly as much as possible, so I thought I'd start with a BehaviorSubject which I can eventually feed with true and combine it with distinctUntilChanged if I ever need to get the full response. This will give me an observable with false initially and that can give me true if I feed that into it:
const fullSubject = new BehaviorSubject<boolean>(false);
Then I've got a function that takes a boolean parameter and returns an observable with the server request (retried, transformed, etc.). As said, the answer can be partial or full, but it can be full even if the input parameter was false at the server's discretion. For example:
interface IdentityData {
...
isFull: boolean;
}
private getSimpleIdentity(full: boolean): Observable<IdentityData> {
return Axios.get(`/api/identity${full?"?full=true":""}`)
.pipe( ... retry logic ...,
... transformation logic ...,
shareReplay(1) );
}
I need to know how can I combine these so that the following is true:
The server needs to be queried at most twice.
If the first answer is a full answer, no further queries must be performed to the server.
If the first answer is a partial answer, and true is fed into fullSubject, a full answer must be requested.
The expected output from all this is an observable that emits either one full response, or a partial response and, when asked, a full response.
Environment: Vue 2.6.11, RxJS 6.5.5, Axios 0.19.2, TypeScript 3.7.5.
Thanks in advance
Here would be my approach:
const fullSubject = new BehaviorSubject(false);
const src$ = fullSubject.pipe(
switchMap(isFull => Axios.get('...')),
take(2), // Server required at most twice
takeWhile(response => !response.isFull, true), // When `isFull`, it will complete & unsubscribe -> no more requests to the server
shareReplay(1),
);
src$.subscribe(() => { /* ... */ });
function getFullAnswer () {
fullSubject.next(true);
}
takeWhile takes a second argument, inclusive. When set to true, when the predicate function evaluates to false(e.g isFull is true) it will send that value as well. –
if I've got it correctly
private getSimpleIdentity(): Observable<IdentityData> {
return fullSubject.pipe(
switchMap(full => Axios.get(`/api/identity${full ? "?full=true" : ""}`)),
shareReplay(1),
);
}
Uses the retryWhen() operator
const source = of("").pipe(map(() => Math.floor(Math.random() * 10 + 1)));
const example = source
.pipe(
tap((val) => console.log("tap", val)),
map((val) => {
//error will be picked up by retryWhen
if (val !== 5) throw val;
return val;
}),
retryWhen((errors) =>
errors.pipe(
tap(() => console.log("--Wait 1 seconds then repeat")),
delay(1000)
)
)
)
.subscribe((val) => console.log("subscription", val));
/*
output:
tap 3
--Wait 1 seconds then repeat
tap 8
--Wait 1 seconds then repeat
tap 1
--Wait 1 seconds then repeat
tap 4
--Wait 1 seconds then repeat
tap 7
--Wait 1 seconds then repeat
tap 5
subscription 5
*/
I'm trying to fit a model using fitDataset(). I can train using the "normal" approach, with a for loop and getting random batches of data (20000 data points).
I'd like to use the fitDataset() and be able to use the entire dataset and not rely on "randomness" of my getBatch function.
I'm getting closer, using the API docs and the example on tfjs-data but, i'm stuck on a probably dumb data manipulation...
So here's how i'm doing it:
const [trainX, trainY] = await bigData
const model = await cnnLSTM // gru performing well
const BATCH_SIZE = 32
const dataSet = flattenDataset(trainX.slice(200), trainY.slice(200))
model.compile({
loss: 'categoricalCrossentropy',
optimizer: tf.train.adam(0.001),
metrics: ['accuracy']
})
await model.fitDataset(dataSet.train.batch(32), {
epochs: C.trainSteps,
validationData: dataSet.validation,
callbacks: {
onBatchEnd: async (batch, logs) => (await tf.nextFrame()),
onEpochEnd: (epoch, logs) => {
let i = epoch + 1
lossValues.push({'epoch': i, 'loss': logs.loss, 'val_loss': logs.val_loss, 'set': 'train'})
accuracyValues.push({'epoch': i, 'accuracy': logs.acc, 'val_accuracy': logs.val_acc, 'set': 'train'})
// await md `${await plotLosses(train.lossValues)} ${await plotAccuracy(train.accuracyValues)}`
}
}
})
here's my interpretation of the dataset creation:
flattenDataset = (features, labels, split = 0.35) => {
return tf.tidy(() => {
let slice =features.length - Math.floor(features.length * split)
const featuresTrain = features.slice(0, slice)
const featuresVal = features.slice(slice)
const labelsTrain = labels.slice(0, slice)
const labelsVal = labels.slice(slice)
const data = {
train: tf.data.array(featuresTrain, labelsTrain),
validation: tf.data.array(featuresVal, labelsVal)
}
return data
})
}
I'm getting an error:
Error: Dataset iterator for fitDataset() is expected to generate an Array of length 2: `[xs, ys]`, but instead generates Tensor
[[0.4106583, 0.5408, 0.4885066, 0.9021732, 0.1278526],
[0.3711334, 0.5141, 0.4848816, 0.9021571, 0.2688071],
[0.4336613, 0.5747, 0.4822159, 0.9021728, 0.3694479],
...,
[0.4123166, 0.4553, 0.478438 , 0.9020132, 0.8797594],
[0.3963479, 0.3714, 0.4871198, 0.901996 , 0.7170534],
[0.4832076, 0.3557, 0.4892016, 0.9019232, 0.9999322]],Tensor
[[0.3711334, 0.5141, 0.4848816, 0.9021571, 0.2688071],
[0.4336613, 0.5747, 0.4822159, 0.9021728, 0.3694479],
[0.4140858, 0.5985, 0.4789927, 0.9022084, 0.1912155],
...,
The input data is 6 timesteps with 5 dimensions and the labels are just one-hot encoded classes [0,0,1], [0,1,0] and [1, 0, 0]. I guess the flattenDataset() is not sending the data in the correct way.
Does data.train needs to output for each data point [6 timesteps with 5 dims, label] ? I get this error when i tried that:
Error: The feature data generated by the dataset lacks the required input key 'conv1d_Conv1D5_input'.
Could really use some pro insight...
--------------------
Edit #1:
I feel i'm close to an answer.
const X = tf.data.array(trainX.slice(0, 100))//.map(x => x)
const Y = tf.data.array(trainY.slice(0, 100))//.map(x => x)
const zip = tf.data.zip([X, Y])
const dataSet = {
train: zip
}
dataSet.train.forEach(x => console.log(x))
With this i get on the console:
[Array(6), Array(3)]
[Array(6), Array(3)]
[Array(6), Array(3)]
...
[Array(6), Array(3)]
[Array(6), Array(3)]
but the fitDataset is giving me: Error: The feature data generated by the dataset lacks the required input key 'conv1d_Conv1D5_input'.
my model look like this:
const model = tf.sequential()
model.add(tf.layers.conv1d({
inputShape: [6, 5],
kernelSize: (3),
filters: 64,
strides: 1,
padding: 'same',
activation: 'elu',
kernelInitializer: 'varianceScaling',
}))
model.add(tf.layers.maxPooling1d({poolSize: (2)}))
model.add(tf.layers.conv1d({
kernelSize: (1),
filters: 64,
strides: 1,
padding: 'same',
activation: 'elu'
}))
model.add(tf.layers.maxPooling1d({poolSize: (2)}))
model.add(tf.layers.lstm({
units: 18,
activation: 'elu'
}))
model.add(tf.layers.dense({units: 3, activation: 'softmax'}))
model.compile({
loss: 'categoricalCrossentropy',
optimizer: tf.train.adam(0.001),
metrics: ['accuracy']
})
return model
What is wrong here?
What model.fitDataset expects are a Dataset, each element inside this dataset is a tuple of two items, [feature, label].
So in your case, you need to create featureDataset and labelDataset, then merge then with tf.data.zip to create trainDataset. Same for validation dataset.
Solved it
so after a lot of trial an error i found a way to make it work.
So, i had an input shape of [6, 5], meaning an array with 6 arrays of 5 floats each.
[[[0.3467378, 0.3737, 0.4781905, 0.90665, 0.68142351],
[0.44003019602788285, 0.3106, 0.4864576, 0.90193448, 0.5841830879700972],
[0.30672944860847245, 0.3404, 0.490295674, 0.90720676, 0.8331748581920732],
[0.37475716007758336, 0.265, 0.4847249, 0.902056932, 0.6611207914113887],
[0.5639427928616854, 0.2423002, 0.483168235, 0.9020202294447865, 0.82823],
[0.41581425627336555, 0.4086, 0.4721923, 0.902094287, 0.914699]], ... 20k more]
What i did was to flatten the array becoming an array of 5 dimensions arrays. Then applied the .batch(6) to it.
const BATCH_SIZE = 20 //batch size fed to the NN
const X = tf.data.array([].concat(...trainX)).batch(6).batch(BATCH_SIZE)
const Y = tf.data.array(trainY).batch(BATCH_SIZE)
const zip = tf.data.zip([X, Y])
const dataSet = {
train: zip
}
Hope it can help others on complex data!!
Given this array in Ruby:
myarray = [name: "John", age: 35]
How do I refer to the age?
I tried myarray[:age] but got an error can't convert Symbol into Integer
Update:
I was trying to simplify my question by extracting what I thought my problem is. I may not understand completely.
I'm experimenting with Dashing and trying to send a number to a meter widget. I've created a variable, 'response_raw' and am trying to send it in the third send event. Here's my code:
SCHEDULER.every '1m', :first_in => 0 do
# Get checks
url = "https://#{CGI::escape user}:#{CGI::escape password}#api.pingdom.com/api/2.0/checks"
`enter code here`response = RestClient.get(url, {"App-Key" => api_key})
response = JSON.parse(response.body, :symbolize_names => true)
if response[:checks]
checks = response[:checks].map { |check|
if check[:status] == 'up'
state = 'up'
last_response_time = "#{check[:lastresponsetime]}ms"
response_raw = check[:lastresponsetime]
else
state = 'down'
last_response_time = "DOWN"
response_raw = 0
end
{ name: check[:name], state: state, lastRepsonseTime: last_response_time, pt: response_raw }
}
else
checks = [name: "pingdom", state: "down", lastRepsonseTime: "-", pt: 0]
end
checks.sort_by { |check| check['name'] }
send_event('pingdom', { checks: checks })
send_event('pingdom-meter', { value: checks[:pt] })
end
In CoffeeScript [name: "John", age: 35] is an array containing single object with two properties (name and age).
Here is how it'll look in plain JavaScript:
myarray = [
{
name: "John",
age: 35
}
];
So, answering your question, to access an age you should take the first element of an array and then reference an age property:
myarray[0].age
or
myarray[0]['age']
But, judging from your question, your're probably using wrong data structure. Why don't you want to use a plain object instead of an array?
person = name: "John", age: 35
console.log "#{person.name}'s age is #{person.age}"
Update
It looks like your question is actually about Ruby and not about CoffeeScript. Though, my answer will remain the same.
To access an age you should take the first element of an array and then reference an age property:
myarray[0][:age]
Since myarray is an array, Ruby expects an integer index, but you're giving it symbol :age instead.
I finally figured it out with Leonid's help. Thank you.
I changed:
send_event('pingdom-meter', { value: checks[:pt] })
to
send_event('pingdom-meter', { value: checks[0][:pt] })
I'm working with a bit of old VB6 code that goes thus...
Dim STATUS As Integer
STATUS = -1
If (Not STATUS) Then
' do something
Else
' do something else
End If
so I was, naturally, wondering which branch of this code is executed. So does anyone know what the numeric values of True and False are in VB6?
True is stored as -1 and false as 0. Any non-zero value is considered as true.
To see why it is so please check - http://www.vbforums.com/showthread.php?t=405047
In VB 6, True has a numeric value of -1. False has a numeric value of 0.
The reason for this is because the Boolean data type is stored as a 16-bit signed integer. Therefore,-1 evaluates to 16 1s in binary (1111111111111111). False is 16 0s (0000000000000000). This produces the relationship that has held throughout the evolution of BASIC: True = Not False.
Not really an answer, but just poking about, I typed this into the immediate window, with these results:
For x = -5 To 5 : ? x, CBool(x), ( x = True ), ( x = False ) : Next x
-5 True False False
-4 True False False
-3 True False False
-2 True False False
-1 True True False
0 False False True
1 True False False
2 True False False
3 True False False
4 True False False
5 True False False
(I tested more values, but only -1 and 0 had anything "interesting" going on. The rest were all True/False/False.) So, empirically, I'd say that the comparison is being done arithmetically unless you cast with CBool. Why? I can't really say...